The iterative cycling of mutation and selection utilized by vertebrate immune systems is a powerful means for generating proteins with diverse properties. In order to apply this approach to the generation of LAGLIDADG homing endonucleases (LHEs) with novel DNA binding and cleavage specificities, we have developed methods to express LHEs as fusion proteins on the surface of cultured B-cells. The surface expressed fusion proteins allow the rapid assessment of the specific binding and cleavage properties of the LHE using flow cytometry, and also can be used to separate populations of cells expressing LHE's with different specificities. Based on these data, we propose the following Specific Aims: In Specific Aim 1, we will generate and characterize new surface expressed LHE scaffolds based on novel LHE's generated by Component 2 (Monnat) and Component 3 (Baker). In Specific Aim 2, we will integrate surface expressed LHE's into immunoglobulin loci of the DT40 cell line such that they become susceptible to the endogenous somatic hypermutation mechanism(s) operating in DT40 cells. We will then use these LHEhypermutating lines to execute two strategies of iterative mutation/selection to identify novel LHE's with desired binding and cleavage properties. In the first, we will use LHE's from Component 3 (Baker) with pre-optimized DNA/protein interfaces towards target sites, and will attempt to directly select LHE variants able to bind and cleave the predicted target. In the 2nd, we will attempt a base pair-by-base pair migration strategy, beginning with presently available surface expressed LHE scaffolds. In Specific Aim 3, we will work on further refining and enhancing our methods and technologies for iterative mutation/selection of LHE variants. In one part of this aim, we will work with Component 2 (Monnat) and Component 5 (Stoddard) on developing an increased sensitivity cleavage assay based on quantum dot nanosensors, for use in both flow cytometry and soluble LHE assays. In the second, we will utilize overexpression of proteins involved in somatic hypermutation to incrase the rate of hypermutation and enhance the spectrum of mutations to include a higher rate of insertions and deletions. The output of this aim directly feeds back to the NGEC LHE design cycle as well, as variants identified here will be passed on to Component 2 (Monnat) and Component 5 (Stoddard) for biochemical and biophysical analysis, with information thus derived incorporated into PSSM matrices for identifying the best engineerable sites by Component 2 (Monnat), and into computational design algorithms developed by Component 3 (Baker).